In planning community services for people with severe and persistent mental disorders, state mental health authorities (SMHA) need to know whether or not companionship programs are effective in facilitating community integration, in improving quality of life, in improving clients' self-esteem, and/or in reducing the need for high cost services such as emergency services. During the past fifteen years the range, variety and sophistication of programs to support seriously mentally ill persons in the community has exploded, enabling more and more people to live in the community rather than in institutions. Companionship programs represent an effort by the mental health system to address the isolation and stigmatization that these clients experience living in the community. Companionship programs are intuitively compelling in that social acceptance and social integration seem most likely to be achieved with the assistance of peer members of the social structure rather than by paid professionals or paraprofessionals. They may also be more cost-effective. Although companionship programs are proliferating throughout the country, there has been little empirical verification of the impact on client outcomes. The proposed project will evaluate client outcomes, as described above, with the randomization of clients to one of three treatment groups: community-at-large companion, mentally ill companion, or no companion. The companions will have a clearly defined task of providing opportunities for clients to meet and connect with others in a wide variety of social situations, and to facilitate a normalized use of leisure time. Small stipends will be provided to all three treatment groups to pay for recreational expenses, and companions will receive a moderate hourly wage for up to four hours of time spent with clients each week for nine months. The activities of clients and companions will be recorded weekly. Outcomes will be assessed three times, at the beginning of the program, at the end of the program, and at the end of a six-month follow-up period. Using a repeated measures design, all groups will be compared on the outcome measures.